Gemini Deep Research And Gems Go Free, 2.0 Flash Thinking Experimental Upgraded

After a series of developer announcements yesterday, Google is rolling out a handful of Gemini app updates for end users centered around an upgraded version of 2.0 Flash Thinking Experimental, which remains in preview.

For starters, more Gemini Advanced features are coming to free users. This follows Google bringing document upload and generating images of people to everyone last month.

Gems are custom versions of Gemini tailored to particular tasks with specific instructions that you provide. Free users (over 18) can now use premade Gems (Brainstormer, Career guide, Coding partner, Learning coach, and Writing editor), or create their own, like a translator, meal planner, or math coach. You can access them on mobile, but the ability to create new ones is only available on the web. The “Gems manager” lets you write instructions, upload files it should reference, and assign a name.

Google is bringing Deep Research to all Gemini (free) users after launching for Advanced subscribers in December. It’s available in over 45 languages. Notably, the research tool is now powered by a new version of Gemini 2.0 Flash Thinking Experimental (more on that below). This model update (from 1.5 Pro) results in improvements across planning, searching, reasoning, analyzing, and reporting. The Thinking aspect of this upgrade means Gemini “shows its thoughts while it browses the web.” Besides opening the model picker, a new chip in the prompt bar will let you Deep Research.

Zooming out, Google is updating 2.0 Flash Thinking Experimental to offer “better efficiency and speed.” The reasoning model is “trained to break down prompts into a series of steps to strengthen its reasoning capabilities and deliver better responses.”

You can now upload files when using the model, with Gemini Advanced users getting a 1 million token context window. Additionally, 2.0 Flash Thinking Experimental Model can now access Gemini apps (previously called Extensions) like Gmail, Google Calendar, Drive, Messages, YouTube, etc. This allows for prompts like: “Look up an easy cookie recipe on YouTube, add the ingredients to my shopping list and find me grocery stores that are still open nearby.”

With this thinking model, Gemini can better tackle complex requests like prompts that involve multiple apps, because the new model can better reason over the overall request, break it down into distinct steps, and assess its own progress as it goes.

A Google Photos app will be available in the coming weeks to deliver “Ask Photos” functionality:

For example, Gemini can look at photos from your recent trip and create a travel itinerary based on the places you visited, or you can ask Gemini to recall information – like when your driver’s license expires.

Google is also using 2.0 Flash Thinking Experimental to power a new ”Personalization (experimental)” model. Gemini “connects with your Google apps and services, beginning with Search, to deliver responses that are even more tailored to your individual needs.”

For example, you can ask Gemini for restaurant recommendations and it will reference your recent food-related searches, or ask for travel advice and Gemini will respond based on destinations you’ve previously searched.

This will expand to Google Photos, YouTube, and other apps in the “coming months” to “provide more personalized insights, drawing from a broader understanding of your activities and preferences to deliver responses that truly resonate with you.”

After selecting this model, Gemini “will analyze [your prompt] and determine if your Search history can enhance the response.” This can be useful for brainstorming and personalized recommendations: